Karen M. Love, Christian Strohmenger
Multivariate regression equations significantly predict the three-dimensional porosity and permeability distribution from regional to field scale in the Zechstein 2 Carbonate. The porosity and permeability distributions relate strongly to facies and matrix mineralogy. Facies control porosity and permeability due to both depositional and diagenetic factors, and matrix mineralogy serves as a control because reservoir dolomites are generally porous whereas calcitized dolomites (dedolomites) are generally non-porous.
Equations first were developed to predict mineralogy and then porosity and permeability using multiple regression analysis with a large, integrated database. Location variables (latitude, longitude, and depth), combined with facies variables, provide statistically significant predictions of mineralogy as well as porosity and permeability. The addition of transformed location variables (squares, reciprocals, logarithms) to the regression models improves the fit by accommodating the complexity of the porosity and permeability distribution. Improved predictions also can be made by areally subdividing the reservoir and generating separate prediction equations for each subdivision. In addition to the porosity and permeability predictions, confidence intervals on the predictions for each subarea can be calculated, providing important guidelines for predicting rese voir quality from one part of the reservoir to another. This information supplements data used for planning well locations, and also suggests optimal areas for three-dimensional seismic surveys.
AAPG Search and Discovery Article #91020©1995 AAPG Annual Convention, Houston, Texas, May 5-8, 1995